profile
viewpoint

aristocratos/bashtop 7205

Linux/OSX/FreeBSD resource monitor

aristocratos/bpytop 165

Linux/OSX/FreeBSD resource monitor

aristocratos/spdtest.sh 72

Internet speed reability tester

issue closedaristocratos/bpytop

[BUG] rxvt-unicode compatibility seems broken

Describe the bug

rxvt-unicode display is quite broken. Also, mouse support seems to be kind of defect.

To Reproduce

  • Install rxvt-unicode ( 9.22-1+b1 here)
  • run bpytop.py

Expected behavior

  • Normal display
  • mouse handling should work

Screenshots

Bildschirmfoto_2020-08-03_17-33-10

Info (please complete the following information):

  • bpytop version: 1.0.0
  • psutil version: 5.7.2
  • (Linux) Linux distribution and version: Debian Stretch with a few buster/sid/experimental upgrades
  • Terminal used: rxvt-unicode
  • Font used: Ormaxx
  • Python version, python3 --version (version 3.6 or above is required): 3.7.2+

Additional context

May be related to additional xressources (although i can't see any related)

closed time in 6 hours

ri0t

issue closedaristocratos/bpytop

"rawdict = _psplatform.disk_io_counters()" raises ValueError "not sure how to interpret line"

Upon startup, i see the intro screen (r'ly?) and it quickly bails out with:

03/08/20 (16:21:32) | ERROR: Data collection thread failed with exception: not sure how to interpret line ' 8 0 sda 61540896 14301543 2070222521 18032078 107627996 33672188 4122914856 94277324 0 74681256 32698424 0 0 0 0\n' Traceback (most recent call last): File "./bpytop.py", line 2315, in _runner collector._collect() File "./bpytop.py", line 2578, in _collect io_counters = psutil.disk_io_counters(perdisk=True if SYSTEM == "Linux" else False, nowrap=True) File "/usr/lib/python3/dist-packages/psutil/init.py", line 2013, in disk_io_counters rawdict = _psplatform.disk_io_counters() File "/usr/lib/python3/dist-packages/psutil/_pslinux.py", line 1081, in disk_io_counters raise ValueError("not sure how to interpret line %r" % line) ValueError: not sure how to interpret line ' 8 0 sda 61540896 14301543 2070222521 18032078 107627996 33672188 4122914856 94277324 0 74681256 32698424 0 0 0 0\n' 03/08/20 (16:21:32) | WARNING: Exiting with errorcode (1). Runtime 0:00:03

Need any more data?

closed time in 6 hours

ri0t

issue closedaristocratos/bpytop

ValueError: 'CPU' is not in list

Hello there,

I installed bpytop as described. 'psutil' is installed. But it keeps failing with this ValueError.

[nico@void ~]$ python --version
Python 3.8.4
[nico@void ~]$ pip --version
pip 20.2 from /usr/lib/python3.8/site-packages/pip (python 3.8)
[nico@void ~]$ lsb_release -a
LSB Version:	1.0
Distributor ID:	VoidLinux
Description:	Void Linux
Release:	rolling
Codename:	void
[nico@void ~]$ python3 -m pip install psutil
Defaulting to user installation because normal site-packages is not writeable
Requirement already satisfied: psutil in /usr/lib/python3.8/site-packages (5.7.2)
[nico@void ~]$ git clone https://github.com/aristocratos/bpytop.git
Klone nach 'bpytop' ...
remote: Enumerating objects: 103, done.
remote: Counting objects: 100% (103/103), done.
remote: Compressing objects: 100% (70/70), done.
remote: Total 103 (delta 52), reused 77 (delta 29), pack-reused 0
Empfange Objekte: 100% (103/103), 638.70 KiB | 587.00 KiB/s, Fertig.
Löse Unterschiede auf: 100% (52/52), Fertig.
[nico@void ~]$ cd bpytop
[nico@void bpytop]$ sudo make install
**[nico@void bpytop]$ bpytop
Traceback (most recent call last):
  File "/usr/local/bin/bpytop", line 3968, in <module>
    CPU_NAME: str = get_cpu_name()
  File "/usr/local/bin/bpytop", line 3710, in get_cpu_name
    name = nlist[nlist.index("CPU")+1]
ValueError: 'CPU' is not in list**

[nico@void ~]$ bpytop -v
bpytop version: 1.0.0
psutil version: 5.7.2

closed time in 6 hours

NicoWde

issue commentaristocratos/bpytop

ValueError: 'CPU' is not in list

Huh, think the cpu name should show up as "Intel Xeon" now anyways...

NicoWde

comment created time in 6 hours

issue commentaristocratos/bpytop

[BUG] Incorrect sorting by Mem% usage and maybe CPU usage

Noticed I had string conversions on both memory and threads in the precompiled sorting expressions.

I'm not sure why I put them there though, but doesn't seem to break anything...

Fixed in v1.0.3

ergoz

comment created time in 6 hours

issue commentaristocratos/bpytop

[BUG] tree and info on a proc crashes

v1.0.3 released

bgkillas

comment created time in 6 hours

issue commentaristocratos/bpytop

Bring back shiftless control.

Changed in v1.0.3

drazil100

comment created time in 6 hours

push eventaristocratos/bpytop

aristocratos

commit sha cf2f59978ea5de72690c322dfcc65916dcb4303f

Fixes for crash on detailed info when showing tree, incorrect sorting for memory, unsupported osx psutil values and hotkey changes.

view details

aristocratos

commit sha 7847b0a565af9ad4fbfca9a3ab458aa8f11f76ff

Updated hotkeys in help

view details

aristocratos

commit sha 28bb245ade55c2431f5afe595f9601b2db71c37d

v1.0.3 Bug fixes and hotkey changes

view details

push time in 6 hours

created tagaristocratos/bpytop

tagv1.0.3

Linux/OSX/FreeBSD resource monitor

created time in 6 hours

issue commentaristocratos/bpytop

[REQUEST] Memory usage in megabytes instead of percents

On a modern cpu it would probably not be too noticeable.

But the process iteration is what takes up maybe 50-75% of bpytops cpu usage, so adding a extra function call for every iteration will most likely have some effect, depending on the number of processes running.

But gonna try it out when I've got some time :)

avently

comment created time in 10 hours

issue commentaristocratos/bpytop

[REQUEST] Memory usage in megabytes instead of percents

This is doable, but will use more cpu since the memory value I can get from psutil is in bytes and would have to be converted with my floating_humanizer() function to convert to KB, MB, GB etc (to not take up too much space). So would add an extra function call for every process on every update.

avently

comment created time in 10 hours

issue commentaristocratos/bpytop

[REQUEST] Remove min width and height requirements

(in what unit this size is?) Characters

the main goal is to remove limitation of min width/height. This isn't possible without creating a lot of issues and possibly crashing because of the way I've coded it. And would require a lot of extra checks on all draw functions to not break.

I'm unable to make terminal workspace with width less than 80 and height less than 24 What terminal/font size are you using where you can't have more than 80 characters in width?

Can look in to it when I've got some time, but not a priority right now.

avently

comment created time in 10 hours

issue commentaristocratos/bpytop

[REQUEST] Remember and restore position after selecting a process

Oh wait, I just realized what you meant. Save the and restore the position is quite doable :)

avently

comment created time in 10 hours

issue commentaristocratos/bpytop

[REQUEST] Remember and restore position after selecting a process

This is intentional to give the details box focus since the "terminate", "kill", "interrupt" commands isn't active for the detailed process if selection is greater than 0. For quick change between which process to show details for you can double click any process in the list.

But I would like to hear some peoples opinion if this is a wanted change.

avently

comment created time in 10 hours

issue commentaristocratos/bpytop

[BUG] Error: Permission denifed: '/proc/partitions'

I'm not the maintainer for the snap package.

See https://github.com/kz6fittycent/bpytop-snap

But I'm gonna guess you didn't run the snap connect commands to give bpytop permissions?

GalGreenfield

comment created time in 11 hours

issue commentaristocratos/bpytop

Bring back shiftless control.

Good feedback! And I and can agree that adding shift modifier wasn't the best idea. It was a last minute change before release.

I don't see why colors and gradient need a shortcut to toggle.

Yeah, did get a bit cluttered, will probably be enough to have them in the options menu.

Also I think that menu and mini should have swaped hot keys. Menu should have capital 'M' and mini version of the app should have small 'm',

Will do.

By the way can we all just appreciate the irony that "mini" got the bigger M xD

Heh, not quite sure what I was thinking there, gonna blame lack of sleep :)

I can probably work in customizable hotkeys at some later point when the most urgent bugs has been fixed.

drazil100

comment created time in 11 hours

issue commentaristocratos/bpytop

[BUG] Memory bar does not account for 100%

Yeah, ubuntu mono can be weird with the sizing of certain symbols. I mostly use Envy Code R patched with Nerd font symbols. But any font that follows the specs in https://github.com/aristocratos/bpytop#prerequisites should work.

drazil100

comment created time in 11 hours

issue closedaristocratos/bpytop

Crash on biew process info

Crash on biew process info

Bpytop exited with errorcode (1).

To Reproduce

v1.0.2 from git, busy system, actively walk on processes, press Enter

Expected behavior

View process info

Info: bpytop version: 1.0.2 psutil version: 5.7.2 mate-terminal 1.16.3 + Anka/Coder Condensed Ubuntu 16.04.6 LTS Python 3.7.6 (built from source)

Additional context

04/08/20 (10:11:35) | ERROR: Data collection thread failed with exception: 'cmd'
Traceback (most recent call last):
  File "bpytop.py", line 2319, in _runner
    collector._collect()
  File "bpytop.py", line 2934, in _collect
    cls.details["cmdline"] = cls.processes[c_pid]["cmd"]
KeyError: 'cmd'
04/08/20 (10:11:37) | WARNING: Exiting with errorcode (1). Runtime 0:10:38

closed time in 11 hours

RomanKharin

issue commentaristocratos/bpytop

Crash on biew process info

Duplicate of #11

Will be fixed in v1.0.3 later today.

RomanKharin

comment created time in 11 hours

issue commentaristocratos/bpytop

[BUG] tree and info on a proc crashes

Will be fixed in v1.0.3 later today.

bgkillas

comment created time in 11 hours

issue commentaristocratos/bpytop

Feature request: Installability as a PyPI package

I wrote this in the readme but I'm gonna repeat it here. If anybody is willing to help with writing tests for bpytop that would be very helpful. And would speed up the process of making it suitable for PyPi packaging.

csachs

comment created time in a day

issue commentaristocratos/bpytop

Feature request: site-wide config file in /etc

Well, it was only 4 lines of code to add :)

gvegidy

comment created time in a day

issue commentaristocratos/bpytop

[BUG] ERROR: list index out of range

Fixed in v1.0.2

AndyLavr

comment created time in a day

issue commentaristocratos/bpytop

Feature request: site-wide config file in /etc

Added in v1.0.2

gvegidy

comment created time in a day

issue commentaristocratos/bpytop

when i scroll up and down it turns into mini mode and then out of mini mode

Added possible fix in v1.0.2

bgkillas

comment created time in a day

issue commentaristocratos/bpytop

[BUG] Cannot run on FreeBSD 12.1-STABLE

Added error handling in v1.0.2

iron-udjin

comment created time in a day

push eventaristocratos/bpytop

aristocratos

commit sha 896828819cd888b50c388531b5d9e7300ada1e01

Added IndexError catch for cpu temperature collection

view details

aristocratos

commit sha ebbc25dc3d61dd0b284b4963d0b950c93ba611a3

Fixed: net_io_counters() not iterating over itself

view details

aristocratos

commit sha be54d84fc50941e6a12574a50ade79e666ab2510

Fix for clearing mouse queue to avoid accidental character interpretation

view details

aristocratos

commit sha 39a94115d3c3ccbb9304734888c3ee6b406dc52a

Added "/etc/bpytop.conf" as default seed for config file creation if it exists.

view details

aristocratos

commit sha 42cab9869ddfe90b52cfc46d31c151f7e9b61224

Added error handling for exception in psutil.cpu_freq()

view details

aristocratos

commit sha 8d3d5ea5dc6b499a6a61071c45914615a5dc4396

v1.0.2 More bugfixes

view details

push time in a day

created tagaristocratos/bpytop

tagv1.0.2

Linux/OSX/FreeBSD resource monitor

created time in a day

issue commentaristocratos/bpytop

[BUG] Cannot run on FreeBSD 12.1-STABLE

I think this should probably be reported to https://github.com/giampaolo/psutil

I can add handling for the exception so it doesn't crash, but can't fix the issue since it's happening in the psutil module.

iron-udjin

comment created time in a day

issue commentaristocratos/bpytop

Feature request: site-wide config file in /etc

Sure can add "/etc/bpytop.conf" as default seed for config file if it exists and no user config file exists yet.

gvegidy

comment created time in a day

issue commentaristocratos/bpytop

Got error at startup: KeyError: 'eth0:0'

No problem :P

pi11

comment created time in a day

issue commentaristocratos/bpytop

Got error at startup: KeyError: 'eth0:0'

Just realized i don't need to iterate over psutil.net_if_addrs() at all. :)

pi11

comment created time in a day

issue commentaristocratos/bpytop

Got error at startup: KeyError: 'eth0:0'

Not quite sure what that "eth0:0" value is, but try new version. Should hopefully not crash now.

pi11

comment created time in a day

issue commentaristocratos/bpytop

[BUG] ERROR: list index out of range

That would overwrite any values fetched from "Package" or "Tdie" with .

I rather go with an IndexError catch, since any value out the Index range would never be used anyways.

Would you mind running python3 -c "import psutil; print(psutil.sensors_temperatures())"

Not quite sure why you are getting uneven number of core temps.

AndyLavr

comment created time in a day

issue commentaristocratos/bpytop

[BUG] rxvt-unicode compatibility seems broken

Will update prerequisites if I find a list of terminals with proper mouse support. But I'm guessing most of the terminals with truecolor support will also have mouse support since they both rely on modern escape sequence standards.

ri0t

comment created time in a day

issue commentaristocratos/bpytop

[BUG] rxvt-unicode compatibility seems broken

See https://github.com/aristocratos/bpytop#prerequisites

Updated with a list of terminals supporting truecolor.

ri0t

comment created time in a day

push eventaristocratos/bpytop

aristocratos

commit sha 37e61a40228c88053f923adf1542136c53bdfd17

Names...

view details

push time in a day

push eventaristocratos/bpytop

aristocratos

commit sha e2d46776559c9df3dd381efe11cf209989a0ce1b

Index update

view details

push time in a day

push eventaristocratos/bpytop

aristocratos

commit sha 6c641ec2c403e512cf3e6b7aab4cbe822d0c4b4a

Added PyPi notice and call for help with setting up testing.

view details

push time in a day

issue commentaristocratos/bpytop

"rawdict = _psplatform.disk_io_counters()" raises ValueError "not sure how to interpret line"

Exception handling added in v1.0.1

ri0t

comment created time in a day

issue commentaristocratos/bpytop

ValueError: 'CPU' is not in list

Fixed in v1.0.1

NicoWde

comment created time in a day

issue commentaristocratos/bpytop

Got error at startup: KeyError: 'eth0:0'

The exception should be handled properly in v1.0.1 just released but still don't know the cause of problem.

pi11

comment created time in a day

created tagaristocratos/bpytop

tagv1.0.1

Linux/OSX/FreeBSD resource monitor

created time in a day

push eventaristocratos/bpytop

aristocratos

commit sha fd8852a160e7dc826e02ca244c283d0e109cf325

Fixed bad assumption of cpu model name string contents.

view details

aristocratos

commit sha d4b101c2d1d356226f4ab521710e4dd9ad94ac06

Added exception catch for psutil io_counters error caused by psutil < 5.7.0 and Linux kernel >= 5

view details

aristocratos

commit sha e436a2a4928298bb3f64d4021ebafd5896aa91d9

added psutil verion requirement

view details

aristocratos

commit sha e4f65e7ef384236efed02cb5c5ef73160146bcfe

Added error handling for psutil.net_io_counters() errors.

view details

aristocratos

commit sha 6a7cd064afc30cc0819c9f6f37b70857d9df70e3

v1.0.1 Bug fixes

view details

push time in a day

issue commentaristocratos/bpytop

"rawdict = _psplatform.disk_io_counters()" raises ValueError "not sure how to interpret line"

Well, the dependencies in the readme does list:

psutil python module (v5.7.0 or later)

But you're right, will add an exception catch and log to error.log since it's not fatal and would only result in disk read/write speeds not being reported.

ri0t

comment created time in a day

issue commentaristocratos/bpytop

Feature request: Installability as a PyPI package

Yeah, probably shouldn't have started at 1.0.0, but I wanted the versioning to be higher than that of bashtop to drive people to bpytop, and save me the issue of bug fixing a 5000 line bash script...

csachs

comment created time in a day

issue commentaristocratos/bpytop

"rawdict = _psplatform.disk_io_counters()" raises ValueError "not sure how to interpret line"

intro screen (r'ly?)

It can be disabled :)

I'm pretty sure you've got an outdated psutil version. You can check which version you got by running bpytop -v, it shows both bpytop version and psutil version.

Psutil versions below 5.7.0 crashes during psutil.disk_io_counters() command if you're running linux kernel 5.

I'm guessing you are running ubuntu/debian or a derivative there of? They likely have an outdated version in the distribution packages.

You can try updating psutil in that case with sudo python3 -m pip install psutil --upgrade.

ri0t

comment created time in a day

issue commentaristocratos/bpytop

ValueError: 'CPU' is not in list

That my mistake assuming a Xeon processor would have "CPU" in it's model name :P

Would you mind running cat /proc/cpuinfo |grep "model name"

Just curious what the actual model name of you cpu is.

Anyways, will be fixed in v1.0.1 later today.

NicoWde

comment created time in a day

issue commentaristocratos/bpytop

Got error at startup: KeyError: 'eth0:0'

Would you mind running:

  1. python3 --version
  2. bpytop -v
  3. python3 -c "import psutil; print('\n'.join(psutil.net_io_counters(pernic=True)))"
  4. python3 -c "import psutil; print('\n'.join(psutil.net_if_addrs()))"
pi11

comment created time in a day

push eventaristocratos/bpytop

aristocratos

commit sha eba317535e523ad08ef2950cb6eb95fb5eff9ae1

added bug report and feature request templates

view details

push time in a day

issue commentaristocratos/bpytop

Feature request: Installability as a PyPI package

Gonna wait a little while so I know it's properly stable. Also gonna need to write some tests. But I have planned to set up automated testing and pypi packaging later on.

csachs

comment created time in a day

PublicEvent

push eventaristocratos/bpytop

aristocratos

commit sha d6577e6c0da20a9c129a7ec30a91f7ec4f58c4be

naming...

view details

push time in 2 days

push eventaristocratos/bpytop

aristocratos

commit sha 6296503821541fa58b3988c40126ec2a92b5e256

tabsizes

view details

aristocratos

commit sha 29095f4d2a0ddba51e4e7c5002c6b0563ee84bbb

Merge branch 'master' of github.com:aristocratos/bpytop

view details

push time in 2 days

push eventaristocratos/bpytop

aristocratos

commit sha 7420290b5b650338d372c8831813168ea0c9ff57

Create FUNDING.yml

view details

push time in 2 days

push eventaristocratos/bashtop

aristocratos

commit sha d2eb76ea5ba9ea98ed673d7bcec3c388b035aedc

Added links to bpytop

view details

push time in 2 days

issue commentaristocratos/bashtop

[REQUEST] Change layout

Well you can turn off the gradient fade on the processes in the options.

Regarding truncating I would say wait for bpytop the python port of bashtop I've been working on the last month. I will be released in about a week. It got better scalability for the cores, full mouse support, uses barely a third of bashtops cpu consumption and has a lot more new features and options.

Right now I'm focusing on finishing bpytop and the only work I've been doing on bashtop is bug fixes.

It's possible I'll work in manual resizing of the boxes in bpytop, but it's a lot of work since I'm not using any outside libraries for the UI, so it's not a priority right now.

turkeyphant

comment created time in 12 days

push eventaristocratos/bashtop

Òscar Casajuana

commit sha ae8d7a9266449b9632f982cfe4fb62c885b1a8a4

Remove extra space

view details

aristocratos

commit sha 89bf185f6b50f846118b1c2bdfe0d6c61ddae4a3

Merge pull request #175 from elboletaire/master Remove extra space in debian prerm script

view details

push time in 14 days

PR merged aristocratos/bashtop

Remove extra space in debian prerm script

My OCD is stirring up my guts with that extra space every time I update packages: _)

image

+1 -1

0 comment

1 changed file

elboletaire

pr closed time in 14 days

issue commentaristocratos/bashtop

[BUG] How is network traffic monitored? Sometimes it doesn't detect anything

You need python 3.6 or greater and psutil module. Can be installed with pyton3 -m pip install psutil.

turkeyphant

comment created time in 15 days

issue commentaristocratos/bashtop

[BUG] How is network traffic monitored? Sometimes it doesn't detect anything

Not really sure how to debug this, since it's not crashing and there is nothing showing in the error log it doesn't point to being an error caused by bashtop. It rather looks like it fails to read /proc/net/dev which would indicate it being locked at times for some reason.

Is there any difference if you switch to psutil data collection?

turkeyphant

comment created time in 15 days

push eventaristocratos/bashtop

aristocratos

commit sha 071cde4a9c2af9a26b8d4b4bddf83852891b7caf

Added info for bug fixing [skip ci]

view details

push time in 16 days

PR merged aristocratos/bashtop

ignore 9p filesystems to fix hang on Crostini

This patch prevents bashtop from hanging when running in a Crostini container. Resolves #173

Crostini (Linux containers on ChromeOS) use the 9P filesystem for communication between various VMs and containers.

+1 -1

1 comment

1 changed file

bolapara

pr closed time in 16 days

issue closedaristocratos/bashtop

[BUG] bashtop hangs at "Running mem collection init" in Crostini container

Describe the bug

When launching bashtop in a Crostini (Linux containers on ChromeOS) container, it hangs at "Running mem collection init" step.

To Reproduce

Launch bashtop in default debian container on ChromeOS

Info (please complete the following information):

  • Bashtop version: v0.9.24
  • (Linux) Linux distribution and version: Debian 10 in Crostini. ChromeOS version 83.0.4103.119
  • (Linux) Data collection type (/proc or psutil): /proc
  • Psutil version: 5.7.2
  • Terminal used:
  • Font used:
  • Bash version, 5.0.3
  • Locales: output of locale -v

LANG=en_US.UTF-8 LANGUAGE= LC_CTYPE="en_US.UTF-8" LC_NUMERIC="en_US.UTF-8" LC_TIME="en_US.UTF-8" LC_COLLATE="en_US.UTF-8" LC_MONETARY="en_US.UTF-8" LC_MESSAGES="en_US.UTF-8" LC_PAPER="en_US.UTF-8" LC_NAME="en_US.UTF-8" LC_ADDRESS="en_US.UTF-8" LC_TELEPHONE="en_US.UTF-8" LC_MEASUREMENT="en_US.UTF-8" LC_IDENTIFICATION="en_US.UTF-8" LC_ALL=

Additional context

Debugging shows it seems to be hanging while trying to set iostat_var, presumably because dev_path is empty.

contents of $HOME/.config/bashtop/error.log

New instance of bashtop version: 0.9.24 Pid: 9640 11:47:37 AM ERROR: On line 2272

closed time in 16 days

bolapara

created tagaristocratos/bashtop

tagv0.9.25

Linux/OSX/FreeBSD resource monitor

created time in 16 days

push eventaristocratos/bashtop

Josh

commit sha faaf3ca047bd78974ea549b3647b3a5b0d9bb679

ignore 9p filesystems to fix hang on Crostini This patch prevents bashtop from hanging when running in a Crostini container. Crostini (Linux containers on ChromeOS) use the 9P filesystem for communication between various VMs and containers.

view details

aristocratos

commit sha c76573ac7dd08cf08fbd576768efb151fcbadb6c

v0.9.25 Fix for crash on linux when 9p filesystem is present

view details

push time in 16 days

issue commentaristocratos/bashtop

[BUG] bashtop hangs at "Running mem collection init" in Crostini container

I would guess psutil is detecting the fileystem as a pseudo fileystem and excluding by default then when passed the all=False flag. So your fix seems like a good choice for /proc and psutil methods to work the same.

I should probably look through the psutil source for more filesystems excluded by default and add them to df excludes.

bolapara

comment created time in 16 days

issue commentaristocratos/bashtop

[BUG] How is network traffic monitored? Sometimes it doesn't detect anything

Are you running multiple network monitor software? Was thinking that it might not get access possibly if something else is constantly reading from /proc/net/dev.

turkeyphant

comment created time in 16 days

issue commentaristocratos/bashtop

[REQUEST] Option to use 24-hour clock

No worries :)

UserBlackBox

comment created time in 16 days

issue commentaristocratos/bashtop

[BUG] How is network traffic monitored? Sometimes it doesn't detect anything

All network info is collected from /proc/net/dev when using linux in both psutil and /proc methods.

You're not running bashtop sandboxed or something like that?

Otherwise i would try tail -f /proc/net/dev while downloading something or running a speed test, to see if there's something going wrong in there.

turkeyphant

comment created time in 16 days

issue commentaristocratos/bashtop

[REQUEST] Option to use 24-hour clock

Huh?

Screenshot at 2020-07-19 23-26-54

UserBlackBox

comment created time in 16 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

The tracing.log could be helpful, but 77M is a bit heavy, maybe if can try to get a couple hundred of lines around where .docker.hyp or com.docker.hyp first appears if you are willing?

trapgate

comment created time in 16 days

issue commentaristocratos/bashtop

[BUG] bashtop hangs at "Running mem collection init" in Crostini container

Just to clarify, does it work as expected when using psutil instead of /proc?

Also what info do you get about the 9p partition in question from df?

Is there a need to have the partition listed with the rest of disks? In that case I could probably find a way of excluding it from the iostat check instead of excluding it from being discovered at all.

bolapara

comment created time in 16 days

pull request commentaristocratos/bashtop

ignore 9p filesystems to fix hang on Crostini

This is a fix for a unreported bug. You'll have to create a bug report with the relevant information and link this pull request.

bolapara

comment created time in 17 days

issue commentaristocratos/bashtop

[BUG] CPU usage not showing as normal user, needs su access

No problem. The default is set to not reversed, but you could have hit the "r" key without noticing which would reverse the order. However that should be indicated by "reversed" being underlined in the processes box titlebar.

freijon

comment created time in 20 days

issue commentaristocratos/bashtop

[BUG] CPU usage not showing as normal user, needs su access

Check your /etc/fstab and see if you have hidepid= under the /proc entry.

Restricting access to PID directories
 
procfs provides the hidepid mount option to restrict access to the /proc/<pid> directories by other users. This is a hardening technique that can make it more difficult for malicious local users to gather information about the processes of other users[2]. 

hidepid=0 | The file located in /proc/<pid>/* will be world readable. This is the default behavior.
hidepid=1 | The /proc/<pid> directories are visible by all users, but users can only access the /proc/<pid> directories they own. This will protect files such as /proc/<pid>/cmdline, which may contain sensitive information.
hidepid=2 | Same as hidepid=1 but the /proc/<pid>  directories of other users are hidden. Since this will hide process  information from all users except root, it may be useful to allow access  to privileged users such as those belonging to the wheel group.
 

freijon

comment created time in 20 days

issue commentaristocratos/bashtop

[BUG] CPU usage not showing as normal user, needs su access

Did you upgrade kernel since it worked like normal? Could possibly be new kernel security features in that case.

freijon

comment created time in 20 days

issue commentaristocratos/bashtop

[BUG] CPU usage not showing as normal user, needs su access

Sounds like an issue with user privileges. Does it work like normal if running sudo bashtop? I know this is an issue on OSX but I haven't seen in on Linux before.

Does the user you are running as have administrator rights?

freijon

comment created time in 20 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

@craigemery Uploaded v0.9.24 which should fix both of your issues.

trapgate

comment created time in 21 days

created tagaristocratos/bashtop

tagv0.9.24

Linux/OSX/FreeBSD resource monitor

created time in 21 days

push eventaristocratos/bashtop

aristocratos

commit sha b8cb8d34b17b4ac3accbd5c8eac16a8a140a0f48

v0.9.24 Fixes for OsX

view details

push time in 21 days

delete tag aristocratos/bashtop

delete tag : v0.9.24

delete time in 21 days

created tagaristocratos/bashtop

tagv0.9.24

Linux/OSX/FreeBSD resource monitor

created time in 21 days

push eventaristocratos/bashtop

aristocratos

commit sha 69232ceb0da079061c472da26c1f4a6b84764109

Fixed psutil script crash on OSX

view details

aristocratos

commit sha 70e844aa261d3d1eefad5f602ff0092b764d4e9a

Fixed error handling for malformed osx-cpu-temp output

view details

aristocratos

commit sha 4f7b332b15f748e691ba26271af7fa4ee58f8910

v0.9.24 Fixes for OsX

view details

push time in 21 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

Haven't found the cause of it, also ran it in a loop for 1 hour+ logging values longer than 8 characters and got nothing.

But changed the script so it ignores any value after the first, which should fix it.

trapgate

comment created time in 21 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

I've added additional error checking that should fix any errors from osx-cpu-temp giving more than 1 value. Will be in next update later today.

trapgate

comment created time in 21 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

Seems to be the same as #166 and is caused by osx-cpu-temp. I haven't found a way to reproduce it. Would you mind running osx-cpu-temp a couple of times and post the output?

trapgate

comment created time in 21 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

It's possible it could happen when running with sudo, but it's far less likely. Anyways, it shouldn't be possible at all in next version :)

trapgate

comment created time in 21 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

@craigemery Your issue is unrelated. It's happens because psutil requires superuser on OSX to get info for non user owned processes, and when it doesn't have privileges all values defaults to a float value I defined in the script.

I've added a cleanup if the default value is set, so it should be fixed in next update.

For now you can run sudo bashtop and it should work fine.

trapgate

comment created time in 21 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

Thanks :) I will leave this open since it's not solved and if others are having the same problem maybe we can get some additional information to fix the issue.

trapgate

comment created time in 22 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

The errors you are seeing is because the python script running psutil aren't giving the expected output when listing processes, but it doesn't seem to cause an exception in the script, otherwise the exception would be logged in the error.log file. So it's really hard to debug exactly what's going wrong. Especially when there is a delay before the crash.

I would suggest for now that you run without using psutil, I will have the pure python version of bashtop (bpytop) ready in a couple of weeks which will be a lot faster and should have a lot robuster error handling.

trapgate

comment created time in 22 days

issue commentaristocratos/bashtop

[BUG] Bashtop crashes after a short time

Does this also happen if you change use_psutil to false? Could you also check which psutil version you have python3 -c "import psutil; print(psutil.version_info)".

Most of the errors seems to be in the graph creation functions which are getting passed the processes box titles "Pid: Program: .." instead of an integer. I could add an integer check in the graph function but would rather not since it would slow down a lot.

You seem to be getting runtime warnings from psutil: /usr/lib/python3/dist-packages/psutil/_pslinux.py:1222: RuntimeWarning: ignoring OSError(61, 'No data available') for file '/sys/class/hwmon/hwmon1/temp1_input' warnings.warn("ignoring %r for file %r" % (err, path),

Which could be the cause of all the problems. Ubuntu haven't updated the psutil version in the distribution packages yet. So a possible fix if you aren't on version 5.7.0 could be running sudo pip3 install psutil --upgrade.

trapgate

comment created time in 23 days

push eventaristocratos/bashtop

aristocratos

commit sha 497041d748719eb08a4c612b5be48bdae707d63e

Added psutil version info field [ci skip]

view details

push time in 23 days

issue commentaristocratos/bashtop

bashtop can't kill selected process mac os

@SmithJson Fixed in v0.9.23

SmithJson

comment created time in 23 days

push eventaristocratos/bashtop

aristocratos

commit sha d8f1541afab1af031870d68722c4f38604beb330

Fixed kill/terminate/interrupt not working in OsX and FreeBSD

view details

aristocratos

commit sha e91e0c790567353dbb77c4d081a945c765331701

Changed Compatibility section to Prerequisites

view details

aristocratos

commit sha 236e0f7dbbddefff74ece92f6364adfdfad2cd2d

v0.9.23 Fixed sending signals to processes in OsX/BSD not working

view details

push time in 23 days

created tagaristocratos/bashtop

tagv0.9.23

Linux/OSX/FreeBSD resource monitor

created time in 23 days

issue commentaristocratos/bashtop

Weird Highlighting Makes it Hard to Read on Mac

Good idea. Could possibly rename the compatibility section to prerequisites and have a subsection for each OS to bring some more attention to it.

parkingthought

comment created time in 23 days

issue commentaristocratos/bashtop

Weird Highlighting Makes it Hard to Read on Mac

No problem mate :) No worries about reporting bugs, I just wish people took a little time reading the readme since many visual problems can be fixed by looking through the Compatibility section.

parkingthought

comment created time in 23 days

issue closedaristocratos/bashtop

[BUG] insecure use of /tmp

Forward of a Debian bug report

Describe the bug

bashtop creates a Python script in /tmp and runs it. But Python adds the directory containing the script to the module search path¹, and /tmp is world-writable, so this in insecure. A local user could plant malicious Python module in /tmp, which would be executed by bashtop.

To Reproduce

Create a psutil.py file with: import os; os.system('(tput reset && cowsay pwned) >/dev/tty; kill -ABRT %s' % os.getppid())

  $ install -m 644 /path/to/psutil.py /tmp
  $ bashtop
   _______
  < pwned >
   -------
          \   ^__^
           \  (oo)\_______
              (__)\       )\/\
                  ||----w |
                  ||     ||
  Aborted

closed time in 24 days

daissi
more